T2

The theorem is mathematically correct for simple linear funnels but breaks down under three conditions that are common in real GTM: (1) power-law concentration (T24) — when a handful of accounts drive most revenue, mean conversion rates are misleading; (2) stage coupling — improving demo conversion might require sales capacity that currently constrains closing; (3) diminishing returns at the bottleneck — the worst stage may be worst because it is structurally unimprovable (e.g., low-intent inbound that will never convert regardless of SDR skill). Critics also note that the theorem says nothing about COST: sometimes it is cheaper to improve a non-bottleneck stage than to fix the bottleneck.

If a funnel has n stages with conversion rates r_1, r_2, ... r_n, then output = Leads * ∏r_i. Because the output is a product of all coefficients, the sensitivity of output to each coefficient is proportional to output / r_i. This means improving the worst (lowest) r_i has the highest multiplicative payoff. Practically: if r_1 (lead-to-meeting) = 5% and r_2 (meeting-to-opportunity) = 50%, then improving r_1 by 1 percentage point (a 20% relative improvement) raises output by 20%, while the same absolute improvement in r_2 raises output by only 2%. The theorem generalizes: always measure the relative improvement (delta_r / r), not the absolute change. In practice this means funnel analysis should identify the stage where conversion is furthest below benchmark, weight by cost-to-fix, and sequence improvement efforts accordingly.

• Goldratt's Theory of Constraints (the foundational industrial analog): throughput of any system is limited by its slowest step; improving non-bottleneck steps yields zero system throughput gain
• Predictable Revenue (Benioff/Ross) field data: SDR-to-AE handoff was the bottleneck for most early SaaS companies; fixing qualification criteria raised closed-won rates by 30-50% without changing demo or closing skills
• McKinsey B2B sales benchmarks: median enterprise funnel has lead-to-close rates of 0.5-2%, with most variance explained by the lead-to-qualified stage (15-30% conversion), not the opportunity-to-close stage (typically 20-30%)
• The mathematical proof is exact: ∂(output)/∂(r_i) = output / r_i, making the worst stage the highest-leverage stage by construction

• Under power-law account concentration (T24), the mean funnel rate is dominated by a few large accounts; optimizing mean conversion ignores the named-account logic that drives most revenue
• Stage coupling means fixing one bottleneck can immediately expose the next — a series of sequential fixes, not a single intervention
• Some bottlenecks are structural: if top-of-funnel intent is low (market is not in-market per T12), no conversion optimization at that stage will overcome the fundamental demand deficit

Run a systematic funnel audit at least quarterly: measure each stage's conversion rate against your company's trailing 4-quarter average and against your segment benchmark. Rank stages by (1-r_i) — the biggest gap from a target rate is your bottleneck. Direct optimization resources there first. Common finding: most companies over-invest in later-stage sales training (closing skills) while under-investing in top-of-funnel qualification criteria, even though qualification has higher multiplicative leverage. Use this theorem to defend budget allocation decisions in quarterly GTM reviews.